Anybody who wants to use UTC can already do that. I would highly recommend it for internal events or appointments. It's a shitty idea to force UTC for domestic times, though.
Problem isn't so much about the total economic output, but rather the distribution of the wealth. Those without a job would also like a piece of the pie.
PU is not more dangerous than the U that it was produced from
These are different materials with totally different properties. You're drawing an over-simplistic conclusion by assuming that they are equally dangerous.
You cannot have it both ways. The shorter the half-life the less time it needs to be stored to become innocuous.
Right, but there's a whole spectrum, not just two ways. Really short or really long half lives are easier to deal with. The worst are the ones in the middle, with half life of 100-1000 years. Short enough that they can produce strong radiation, and long enough that it will remain a problem for generations.
Good luck expressing that in a programming language.
Machine learning is not done by expressing rules in a programming language. It is taught by example. In theory, all you need is collect a bunch of examples, or have a way to correct the AI when it's making a mistake.
You might fool beginner players, you won't fool GM's. Pruning in general is already questionable, as you are deleting possibilities based on assumptions that are in turn based on a limited subset of your data. Early pruning leads to big performance gains but may also easily overlook possibilities because certain paths are assumed wrong
If you get big performance gains that means it plays better overall, so it will be more likely to fool GMs as well.
Keep in mind that every human player also uses early pruning. A GM looking at board typically prunes 25 of the 30 possible moves right away.
TO get a 200x speed up on the test set they must have a 200x slow down on average elsewhere.
Possibly, yes. But typical images only form an absolutely tiny subset of all possible images. If we can speed up recognition of real-life images, at the cost of losing speed in fields of noisy random pixels, it's still a useful improvement.
You all thought 'cutting the cord' was going to save you money and get you what you want and only what you want?
Depends. As the article says, most people only use a couple of streamers. We only have 2 subscriptions at our house, and it's still less than half the price of cable, while it offers more entertainment that I'm interested in. The beauty of entertainment is that it is interchangeable. I don't have to watch a particular movie if I can watch something equally entertaining instead.
Netflix dropped the stuff I really wanted to watch. So I dropped them.
They even drop stuff partially. For some TV show that I wanted to watch, they still offer season 2 and further, but withdrew season 1. It boggles the mind why anyone would think that's a good idea.
Slashdot Mirror
*international* events.
I'm guessing people would rather keep it that way - and most likely on DST (summer) time
Permanent DST sucks for Spain in the winter.
Berliners can start enjoying the morning light at 3 am during mid-summer (sunrise would be at 3:43am without DST).
3 am or 4 am sunrise hardly makes a difference if you want to sleep until 7 or 8 am. Get some good curtains if the light bothers you.
Timezones are pointless, let's all just use UTC.
Anybody who wants to use UTC can already do that. I would highly recommend it for internal events or appointments. It's a shitty idea to force UTC for domestic times, though.
The EU is to big to have on timezone
Right, but they can be told to stay on winter time, which makes the most sense for a large part of the central european timezone.
NASA says glacier is speeding up ? NASA is lying.
NASA says glacier is slowing down ? NASA is lying.
Problem isn't so much about the total economic output, but rather the distribution of the wealth. Those without a job would also like a piece of the pie.
PU is not more dangerous than the U that it was produced from
These are different materials with totally different properties. You're drawing an over-simplistic conclusion by assuming that they are equally dangerous.
We have the technology to permanenly and safely dispose of all radioactive isotopes,
So why isn't it used on a wide scale ?
You cannot have it both ways. The shorter the half-life the less time it needs to be stored to become innocuous.
Right, but there's a whole spectrum, not just two ways. Really short or really long half lives are easier to deal with. The worst are the ones in the middle, with half life of 100-1000 years. Short enough that they can produce strong radiation, and long enough that it will remain a problem for generations.
and had longer half lives than the important isotopes being dumped.
That's not necessarily an improvement. The longer the half life, the less radiation per second.
Good luck expressing that in a programming language.
Machine learning is not done by expressing rules in a programming language. It is taught by example. In theory, all you need is collect a bunch of examples, or have a way to correct the AI when it's making a mistake.
I expect the car to follow the rules of the road, and within those rules try to protect its passengers as much as possible.
You already know you cannot please everyone, so you leave it to the maths. Harm no human unless an equivalent or greater harm comes to 2+ humans.
This would allow for killing of innocent bystanders as long as their organs can be harvested to save 2+ others.
Seriously, how do you define "ethics" so that it would be an acceptable definition to, well, everyone?
Simple answer: just make them follow the law, and, if necessary, change the law when problems are found.
There is no reason that a device that is waterproof to a depth of 6 feet for 30 minutes cannot also have a replaceable battery.
Or, as a compromise, the device could be made waterproof with original battery, but still allow you to update battery while breaking the seals.
He's now gearing up to plunk down more money on another pair....
Sounds like it's working as designed.
You might fool beginner players, you won't fool GM's. Pruning in general is already questionable, as you are deleting possibilities based on assumptions that are in turn based on a limited subset of your data. Early pruning leads to big performance gains but may also easily overlook possibilities because certain paths are assumed wrong
If you get big performance gains that means it plays better overall, so it will be more likely to fool GMs as well.
Keep in mind that every human player also uses early pruning. A GM looking at board typically prunes 25 of the 30 possible moves right away.
TO get a 200x speed up on the test set they must have a 200x slow down on average elsewhere.
Possibly, yes. But typical images only form an absolutely tiny subset of all possible images. If we can speed up recognition of real-life images, at the cost of losing speed in fields of noisy random pixels, it's still a useful improvement.
You all thought 'cutting the cord' was going to save you money and get you what you want and only what you want?
Depends. As the article says, most people only use a couple of streamers. We only have 2 subscriptions at our house, and it's still less than half the price of cable, while it offers more entertainment that I'm interested in. The beauty of entertainment is that it is interchangeable. I don't have to watch a particular movie if I can watch something equally entertaining instead.
Netflix dropped the stuff I really wanted to watch. So I dropped them.
They even drop stuff partially. For some TV show that I wanted to watch, they still offer season 2 and further, but withdrew season 1. It boggles the mind why anyone would think that's a good idea.
You can search the web but "which streaming service is show X on" is a surprisingly difficult query.
It gets even worse if you don't live in the US, because the selections are different.
What we need is a subscription service, that manages your subscription services.
Even better, you need a couple of them.
You can only penetrate human tissue a certain depth at the comm frequency used by these devices, regardless of the signal amplitude
Not really a relevant objection, since the extra distance between normal use vs remote hacker is not going through human tissue.
The manufacturers struggle to get it to work beyond a few centimeters.
I've seen one work about 6 feet away, and it didn't require a large box, or directional antennas.
The wireless has to be turned on by placing a magnet against my skin above the defibrillator.
That's not going to work well when it needs to talk to the bedside monitor every night.